I installed Python3.6, the Anaconda distribution, on two different machines. I cannot swear that I used the same installer file, although I thought. I see the same when I try to check versions of Python, Anaconda and numpy:
I was getting small numerical differences. After some debugging, I succeeded to reduce the problem to numpy.exp calls. Just run the code
import numpy as np
x = -0.1559828702879514361612223
y = np.exp(x)
print("The exponential of %0.25f is %0.25f" % (x, y))
I get
The exponential of -0.1559828702879514361612223 is 0.8555738459791129013609634
on the first ("server") computer and
The exponential of -0.1559828702879514361612223 is 0.8555738459791127903386609
in the second ("local") machine.
I know that a float does not have 25 decimal precision, but these differences propagate in my code and take place around the 12th decimal number.
What could be causing different behaviors?